After Mythos, Anthropic launches Claude Managed Agents to speed up agentic AI development

Claude Managed Agents will make it easier for developers to build and run AI agents. (Representational image made with AI)

Anthropic is expanding its enterprise business tools and has launched a new product called Claude Managed Agents, which will make it easier for developers to build and run AI agents. With the new tool, developers only need to define the tasks, tools, and guardrails for building an AI agent, while Claude Managed Agents handles the remaining parts of the development cycle. This tool will enable AI agents to go from prototype to launch in days rather than months.

With the launch of Claude Managed Agents, the AI company aims to further expand its enterprise business, which is already growing very fast. The company reported annualized recurring revenue of $9 billion at the end of 2025, and it has now reached $30 billion in March 2026.

The majority of the company's growth has come from its Claude Platform, an enterprise product that allows developers to access the company's AI models through an API. Claude Managed Agents is expected to further fuel this growth, as the tool will free developers from building complex infrastructure for AI agents and allow them to focus more on improving the user experience.

Agents built with Managed Agents will not require servers, security systems, or execution environments, as everything will run on Claude's cloud infrastructure. Additionally, it comes with a system that allows agents to work autonomously. They can call tools on their own, manage context, and recover from errors.

The new Managed Agents also includes several special features that allow developers to build agents that can run for hours, create and manage other agents, and access APIs, databases, or internal tools. The tool also provides greater monitoring control to developers, allowing them to see which tools the agent uses, what decisions it makes, and where errors occur.

Anthropic says the tool is currently available in public beta on the Claude Platform starting today. In its testing, the company found that outcomes improved by 10 points compared to standard prompting.

The new tool comes a day after the company launched its latest artificial intelligence model called Claude Mythos Preview. The company claimed Claude Mythos is so powerful that it was able to spot thousands of cybersecurity vulnerabilities that were missed by humans across apps and operating systems.

Anthropic said it is working with over 40 tech companies to use Mythos and fix such vulnerabilities before potentially releasing the model to the public.